All analysis starts with a daily rollup of failed payments. It's the only way to accurately exclude transition periods, in progress campaigns, and outlier days.

Most brands compare retention flows by comparing time period A vs time period B. This is like flipping a coin when it comes to your retention. The key is to look at rolling time periods to surface natural variance and help with attribution.

Once a payment fails, there are 4 outcomes: Customer updates card, retry succeeds, customer cancels, customer reaches the end of the recovery process and passively churns. Many tools completely ignore customer cancels, inflating recovery rate and hiding a key indicator.

Don't just look at recovery rate, zoom into recovery rate via retries alone, and via card updates alone. Also monitor cancellation rates, and passive churn rates. Trying "Smart" retries? Zoom into retry success for attribution.

Every company has natural variance in both their passive and active churn. Metrics go up and down day by day based on whose payment failed. We identify natural variance, which helps avoid attributing an increase or decrease to a strategy change that was simply going to happen anyway.

In cases where natural variance is high, segmentation can be used to zoom into segments with less variance - like customers that didn't use a discount, or customers who've placed several orders. We can also segment by processor, decline code, and just about any other attribute.